Interrupted time series analysis (ITSA) has emerged as a prevalent method for evaluating the effects of policy or broad healthcare interventions over time. Within ITSA, segmental regression analysis offers a robust evaluation of intervention impacts. However, for interventions exhibiting seasonality and autocorrelation, the Autoregressive Integrated Moving Average (ARIMA) model presents a valuable alternative.
This presentation explains the foundational theory of ARIMA models and their application in assessing the efficacy of SurgeCon, a pragmatic emergency department (ED) management platform implemented to reduce wait times and improve patient flow in Newfoundland and Labrador EDs without significant changes to workforce volume and composition. Additionally, we detail the process of model selection, fit assessment, and result interpretation, providing insights into the effectiveness of SurgeCon and the utility of ARIMA in evaluating complex healthcare interventions.